connections to resist seismic stressing

huangyhg

connections to resist seismic stressing
hello structural engineers,
as a welding engineer i read with interest the article by ron hamburger
and others on the new aws d1.8, explaining the seismic welding
supplement, in the february 2007 issue of the welding journal at page 28.
not being a structural engineer and in the absence of helpful
sketches i had some difficulty in visualizing the configurations
that produced brittle fractures.
as one of the requirements for avoiding brittle fractures seams to be
the capability of inducing "ductile plastic hinging in the beams at the
beam-column joint", i asked the opinion of the authors, through aws,
on the configuration proposed hereafter, but somehow i got no answer.
a suitable console is welded on the column. the beam rests on the
console and transversal movements are prevented by proper welded
restraints. longitudinal and bending stresses developing between the column and the beam, are sustained by powerful springs that provide tension keeping the beam in place, without welding the beam to the column. hinging seems to be assured.
i would like to ask your expert opinion if the above configuration
could be considered for providing the required stability even in the
event of seismic events.
thanks,
sort of a giant spring clamp at the connections?
it seems like it should work but the springs would need to be extraodinarily stiff. connecting the springs to the beams would need some thought as well. you'd need a lot of research as to the performance of a set of springs under seismic type loading.  
one of the problems with the spring idea is that the structure needs a relatively stiff joint until yielding (plastic hinge) occurs under the eq loading. then at that point you need yielding with rotation.
the springs only provide a constant stretching under conditions from zero to max. load.
when you get right down to the basic physics, any moment resisting connection is a spring, no matter what the connection, as they are all ei related, the overall system allowing a specific deflection per unit of force.
it's just a matter of the spring constant as to how flexible the joint is.  you could have a variable spring constant too, but it is still a spring system.
mike mccann
mccann engineering
that is exactly why engineers are searching for a better solution!
as mike said the non-seismic moment connections are esentially like springs (similar in rotational characteristics). we do not want failure to occur in these connections since the building might collapse! we want an energy absorbing element away from the connection.
thank you all,
if you think that:
"it seems like it should work [...]"
and
"you'd need a lot of research [...]"
your answers give me a good feeling.
of course the springs ought to be stiff.
but less than the brittle welded structures that crack under load.